When trying to enable safe human-animal interaction, it is vital to be able to interpret the animal's signalling correctly to avoid injury to the personand distress to the animal. However, it has been shown that children and adults often do not understand dogs' body signalling (Reisner & Shofer 2008). Without tuition, children look mainly at the dog's face. In addition, children often confuse a fearful or angry dog with a friendly one (Lakestani, 2006, Meints, Racca & Hickey, 2010). Specific aims: This experimental study will investigate how children from 3-5 and adults perceive and interpret dogs' stress signaling. We will then teach children and parents to link their perception of the dog with the correct interpretation of dogs' stress signalling. This way, we help children to understand dog's stress signalling and interpret dog behaviour better and enable them to interact with dogs more safely in the future. We will also teach parents to recognise and interpret the stress signals correctly, so that they also know how to treat their pet dog (and other dogs) more appropriately - and are in turn able to convey this information to others in future. As a consequence, dogs' well-being will also be significantly enhanced. Implications for the health of children, and dogs: If we can teach children to recognise and interpret their own and other dogs' stress signalling better, both sides will profit significantly: children will get bitten less, and the (family) dog will enjoy mor respectful and more appropriate treatment. With bite figures from interview data as high as 47% (Beck & Jones,1985; Spiegel, 2000), recent National Health Service statistics in the UK showing a 40% increase in dog bite figures (NHS, 2008), and our own recent data showing 26% of the population having been bitten (Meints et al., in prep.), we are addressing a serious and wide-spread - but largely avoidable - problem. Wider implications: We promote safe living together of humans and animals and aim to maximise health and safety for children, adults and animals. Our experimental study will be used to teach and inform dog owners and their families, will be fed back to the general public, inform the media, be usable for organisations like Dogs for the disabled and where dogs are used in therapeutic or educational settings and will be helping to teach the public about dog signalling effectively. Research design and methods: Using an experimental design (cross-sectional with longitudinal) we will test children from 3-5 years (24 per experimental group) on their evaluation of a video of dogs in various situations and degrees of being distressed (e.g. licking nose, turning away, growling) recording eye-movements and scan patterns using a Tobii eye-tracker and asking for and recording verbal evaluations of dog signals in the initial baseline test-phase. We will then teach them the correct interpretation reusing the same videos, explaining the dog's signalling. After this, we will re-test them on novel videos with similar situations. We will re-test 6 and 12 months later. This design has the advantage that we can measure conscious verbal interpretations, and also see where on the dog participants focus. Thus, we can compare utterances with looking behaviour and detect discrepancies should they exist. We will also test children's parents using the same procedure. In addition, we will gather information on bite incidents, parents' knowledge on dog signalling, family status and SES. We can also use the combined information to establish which situations are least well evaluated to help children and parents to interpret dogs' stress signals more correctly in future.